Inhibition of PI3K or Integrin-Linked Kinase (ILK) Target Primary AML Cells within the Bone Marrow Microenvironment in the In Vitro Co-Culture System.

In hematological malignancies, there are reciprocal interactions between leukemic cells and cells of the bone marrow microenvironment such as marrow stromal cells (MSC). It is proposed that specific niches within the bone marrow microenvironment provide a sanctuary for subpopulations of leukemic cells to evade chemotherapy-induced death, and we indeed demonstrated that MSC protect primary AML cells from Ara-C induced apoptosis in vitro (Konopleva, Leukemia 2002). Integrin-linked kinase (ILK) has been shown to directly interact with β integrins and phosphorylate AKT in a PI3-kinase(PI3K)-dependent manner to promote cell survival and proliferation. In this study, we tested the hypothesis that selective inhibition of ILK signaling will provide a novel approach for targeting both leukemic cells and cells in their surrounding microenvironment. Direct co-culture of human MSC and leukemic NB4 cells results in activation of PI3K/ILK/AKT signaling as evidenced by enhanced ILK kinase activity, elevated phospho(p)-Akt, p-GSK3β and nuclear translocation of β-catenin. Both, PI3K inhibitor LY294002 (10μM) and specific ILK inhibitor QLT0267 (10μM) inhibited stroma-induced activation of AKT and suppressed GSK phosphorylation. This resulted in massive induction of apoptosis which was not abrogated by stromal co-culture (AnnexinV positivity %, MSC(−) vs MSC(+); 51.4+2.5 vs 55.8+3.5 p=0.26, LY 47.0+8.1 vs 47.9+6.1 p=0.85, 48hrs). In contrast, MSC co-culture effectively blocked apoptosis induced by MEK inhibitor PD98059 despite activation of pERK (62.5+3.2% vs 45.6+2.3%, p=0.02). We next examined anti-leukemia effects of PI3K and ILK inhibitors in the co-culture system of primary AML and human MSC. AML blasts from 7 primary AML samples with high (>54%) blast count were co-cultured with MSC for 24 hours, after which they were exposed to 10μM LY294002 or QLT0267 for 4–8 days. After this period, induction of apoptosis was analyzed in non-adherent AML cells by Annexin V flow cytometry after gating on the CD90-negative (non-MSC) population. To control for differences in spontaneous apoptosis, we calculated % specific apoptosis as (test - control) x 100 / (100 - control). MSCs protected leukemic blasts from spontaneous apoptosis in all 7 samples studied (mean annexin V positivity, 49.5±7.2% vs 25.3±4.8%, p<0.001). In contrast, inhibition of PI3K/ILK signaling induced unopposed apoptosis even in MSC co-cultures (% specific apoptosis, LY294002, 30.3±4.8%; LY+MSC, 28.3±7.7%; QLT0267, 26.9±9.8%; QLT+MSC, 33.1±9.3%, p>0.3 comparing cell death in the presence or absence of MSC). This resulted in corresponding loss of viability (% of control, LY294002, 66.0±11.0%; LY+MSC, 57.6±11.2%; QLT0267, 66.4±7.28%; QLT+MSC, 50.4±11.3%, p>0.1 comparing viability in the presence or absence of MSC). These observations indicate that disruption of leukemia/stroma interactions by specific PI3K/ILK inhibitors represents a novel therapeutic approach to eradicate leukemia in the bone marrow microenvironment. Further studies are aimed at the elucidation of the role of the BM microenvironment and its ability to activate specific signaling pathways in the pathogenesis of leukemias. Focus on this stroma-leukemia crosstalk may result in the development of strategies that alleviate the acquisition of a chemoresistant phenotype and enhance the efficacy of therapies in hematological malignancies.